https://ntrs.nasa.gov/search.jsp?R=19940032314 2020-06-16T10:53:31+00:00Z

NASA Technical Memorandum 4555

June 1994

Review of Advanced Radiator Technologies for Spacecraft Power Systems and Space Thermal Control

Albert J. Juhasz and George P. Peterson

National Aeronautics and Space Administration NASA Technical Memorandum 4555

1994

Review of Advanced Radiator Technologies for Spacecraft Power Systems and Space Thermal Control

Albert J. Juhasz Lewis Research Center Cleveland, Ohio

and

George P. Peterson Texas A&M University College Station, Texas

National Aeronautics and Space Administration Office of Management Scientific and Technical Information Program Trade names or manufacturers' names are used in this report for identification only. This usage does not constitute an official endorsement, either expressed or implied, by the National Aeronautics and Space Administration. REVIEW OF ADVANCED RADIATOR TECHNOLOGIES FOR SPACECRAFT POWER SYSTEMS AND SPACE THERMAL CONTROL

Albert J. Juhasz Lewis Research Center Cleveland, Ohio

and

George P. Peterson Texas A&M University College Station, Texas Summary Introduction

The thermal management of manned spacecraft traditionally The traditional means for rejecting heat from manned space- has relied primarily on pumped, single-phase liquid systems to craft are heat-rejection systems composed of single-phase fluid collect, transport, and reject heat via single-phase radiators. loops (Peterson 1987). These single-phase fluid loops use a Although these systems have performed with excellent reliabil- mechanically pumped coolant to transfer heat from the habita- ity, evolving space platforms and space-based power systems tion portion of the spacecraft to the radiators where it is rejected will require lighter, more flexible thermal management sys- to the space environment. Although these systems have per- tems because of the long mission duration, large quantities formed with excellent reliability in the past, evolving space of power system cycle reject heat, and variety of payloads platforms and space-based power systems will require more involved. The radiators are critical elements in these thermal flexible thermal management systems because of the multiyear management systems. This report presents a two-part overview mission durations, large quantities of heat to be rejected, long of progress achieved in space radiator technologies during the physical distances, and large variety of payloads and missions eighties and early nineties. Part I contains a review and com- that must be accommodated (Mertesdorf et al. 1987). parison of the innovative heat-rejection system concepts pro- In general, space thermal management systems, whether posed during the past decade, some of which have undergone serving life support or future space power systems, consist of preliminary development to the breadboard demonstration three separate subsystems: stage. Included are space-constructable radiators with heat (1) A heat acquisition subsystem that collects heat from pipes, variable-surface-area radiators, rotating solid radiators, the various payload or power system heat-rejection moving-belt radiators, rotating film radiators, liquid droplet interfaces radiators, Curie point radiators, and rotating bubble-membrane (2) A heat transport subsystem that transports heat from radiators. the acquisition sites to the radiating surfaces Part II contains a summary of a multielement project effort, including focused hardware development under the Civil Space (3) A heat-rejection subsystem composed of radiating sur- Technology Initiative (CSTI) High Capacity Power program. faces that form the space radiator A key project under this program carried out by the NASA An example of a typical space thermal management system Lewis Research Center and its contractors was the develop- proposed for large space platforms, namely a two-phase heat- ment of lightweight space radiators applicable to the Space rejection system consisting of the subsystems listed above, was Exploration Initiative (SEI) power systems technologies. Prin- presented by Edelstein (1987). These three subsystems com- cipal project elements include both contracted and in-house prise a thermal utility that would employ the high latent heat of efforts conducted in a synergistic environment designed to a working fluid to transport heat from its acquisition sources to facilitate accomplishment of project objectives. The contracts the radiators, where it would be rejected by radiation to the with Space Power Inc. (SPI) and Rockwell International are particular space environment. aimed at development of advanced radiator concepts, whereas The last of the three subsystems, the radiators, are critical the in-house work has been guiding and supporting the overall components of virtually all proposed space-borne installations. program with system integration studies, heat pipe testing, In most current designs, the radiator is composed of an array of analytical code development, radiating surface emissivity tubes or tube-fin structures through which liquid coolant is enhancement, and composite materials research to develop circulated. The tube wall must be sufficiently thick to minimize and analyze lightweight, high-conductivity fins. These tasks micrometeoroid penetration. As a result, the radiator mass are key prerequisites in the effort to reduce specific mass of could comprise as much as half of the total system mass (Juhasz space radiators. and Jones 1987). The technical challenges associated with the development of Part I.—Innovative Radiator heat-rejection systems capable of meeting future requirements have been described previously (Ellis 1989). Presented here is Technologies a review and comparison of the heat-rejection systems that have been proposed for development for space platforms and space- Space-Constructable Heat Pipe Radiator based power systems. The first part discusses innovative concepts that have under- The heat-rejection system presently used on the space shuttle gone only limited development but are documented for poten- orbiters consists of over 250 small, parallel tubes embedded tial future consideration. These include space-constructable within a honeycomb structure. Warm, single-phase Freon from radiators, variable-surface-area radiators, rotating solid the heat collection and transport circuit is circulated through radiators, moving radiators, and rotating bubble-membrane these tubes (fig. 1). Heat is transferred from the coolant by radiators. convection to the tube walls, conduction through the honey- The second part contains a summary of a multielement comb structure, and finally, radiation to space. Application of project effort including focused hardware development under this technology to the station heat-rejection subsystem would the CSTI High Capacity Power program carried out by the require over 750 interconnected tubes. Moreover, if only a NASA Lewis Research Center and its contractors for the single redundant loop were used, a puncture in any single tube purpose of lightweight space radiator development in support could disable the entire system, making this type of system of Space Exploration Initiative (SEI) power systems technol- infeasible for long-tetra missions. ogy. Principal project elements include both contracted and in- A space-constructable radiator (SCR), composed of a series house efforts conducted in a synergistic environment designed of individually sealed heat pipe elements similar to that shown to facilitate accomplishment of project objectives. The con- in figure 2, has been proposed (Ellis 1989), and several advan- tracts with Space Power Inc. (SPI) and Rockwell International tages of this type of system over pumped single-phase fluid are aimed at development of advanced radiator concepts, loops have been identified. These advantages include a signifi- whereas the in-house work has been guiding and supporting the cant reduction in weight due to the reduction in fluid inventory, overall program with system integration studies, heat pipe increased heat-rejection capacity due to the uniform tempera- testing, analytical code development, radiating surface emis- ture of the radiating surface, and increased reliability, because sivity enhancement, and composite materials research aimed at penetration by a single micrometeoroid or piece of space debris development and analysis of lightweight, high-conductivity, would result in the failure of only a single heat pipe element and high-emissivity fins. These tasks are considered to be key therefore cause only a slight degradation in performance. prerequisites in the effort to reduce specific mass of space High-capacity, SCR elements have been investigated in radiators. several shuttle flight experiments, including the STS-3 flight

Shuttle-Type State-of-the-Art Space Station High-Capacity Radiators Heat Pipe Radiator Heat Pipe Radiators /--Heat transfer 32 °C 32 °C (90 °F) 4.5 °C (40 °F) / contact surface

4.5 °C (40F °F

• Fluid must be pumped over • Fluid needs to be pumped • Two-phase fluid interfaces radiator entire radiator area only through radiator manifold only at manifold

• Over 250 parallel tubes • Less than 75 heat pipes • Only 50 heat pipes required for required for 25-kW station required for shuttle type system 75-kW station • Puncture of any tube destroys • Heat pipes independent of • Heat pipes independent of each other radiator each other; a puncture of one tube only reduces • Contact heat exchanger allows on- efficiency orbit assembly and repair of radiator Figure 1.— Evolution of heat-rejection in spacecraft. fluid changes phase. Oren (1982) gives the results of an Monogroove heat pipe investigation involving two types of flexible roll-out fin radi- ators in which no working fluid phase change was required. Honeycomb The first radiator had a rolled-up fin with a plastic or elasto- meric tube attached to both sides. When gas pressurization was allowed to inflate the two tubes, the fin unrolled, and provided a substantial increase in the surface area. The second radiator used aluminum radiator tubes that were wound in the form of a helical spring configuration to form a cylinder covered by the tin material. This variable-surface-area radiator, which used the inherent spring force (similar to a jack-in-the-box) for deployment, was intended to meet heat-rejection needs of up to 12 kW (Leach and Cox 1977; Oren 1982). Because no phase change was required, an ethylene glycol/water solution was Figure 2.—Space-constructable radiator panel configuration. proposed as the working fluid. Several types of variable-surface-area radiators that utilize a liquid vapor phase change and the associated increase in vol- of the Thermal Canister (Harwell 1983), the STS-8 Heat Pipe ume have been proposed. Figure 4 illustrates the proposed Radiator Experiment (Alario 1984), and the recent Space operation of these radiators. As shown, in their simplest form, Station Heat Pipe Advanced Radiator Element (SHARE) flight these types of radiators are composed of two thin-walled sheets test (Rankin, Ungar, and Glenn 1989; Kossan, Brown, and sealed along the edges and formed into a concentric roll. This Ungar 1990). Results of these three flight tests, along with those roll extends, or rolls out, because of the increased vapor pres- of numerous ground tests, have demonstrated that heat pipe sure generated by heating the wick structure in the evaporator. radiators present a feasible alternative to pumped single-phase This wicking structure can, in some cases, line the inside of the systems. entire fin to assist in liquid return. Once the vapor condenses, In addition to the space station, such space-constructable the fin curls back or retracts to the stowed position. heat pipe radiator systems could be utilized in solar dynamic (SD) power systems (Brandhorst, Juhasz, and Jones 1986; Gustafson and Carlson 1987). In this application, the radiators must reject the nonconvertible thermal energy portion of the total heatenergy supplied to the power system. They thus repre- sent a critical component of the overall development of space- based power systems. Figure 3 shows a typical SD power module design that incorporates a space-constructable heat R pipe radiator system. As illustrated, the radiator could be pZ segmented into several panels for redundancy, with a few aE excess panels incorporated to serve as backup spares for panels with degraded performance. Although these space-constructable heat pipe radiators have performed adequately under realistic thermal/vacuum test con- ditions and during several shuttle flight tests, the application of this technology to an SD power system for the space station, for example, would require about 50 heat pipes, each 10 to 15 m long, to reject the 75 kW required by the 1989 space station design (Ellis 1989).

Variable-Surface-Area Radiator

The concept of a flexible, variable-surface-area radiator that can absorb high peak heat loads for brief time intervals was first introduced in 1978 (Leach and Cox 1977). These types of radiators can be classified into two major categories: (1) those i .'—....nyci ...c"Iuiy in which no phase change occurs and (2) those in which the Figure 3.—Solar dynamic power module. Section A-A Heat in Ev wi sti (a) Heat in Screen wick

Metal foils seam-welded along the edges (b) t t Figure 4.—Roll -out fin tubular segment...... t ...... t t

Liquid E?- Figure 5 illustrates the principle of operation for this type of radiator. Initially, the working fluid within the fin exists as a subcooled or saturated liquid (fig. 5(a)). Because of the flexibil- Heat radiates ity of the fin, heat addition and rejection occur at constant (c) pressure. Heat added to the evaporator vaporizes the working fluid, which expands, thereby causing the fin to extend (fig. 5(b)). This permits the entire external surface of the fin to radiate heat to space (fig. 5(c)). As the vapor condenses, the longitudinal stiffness causes the fin to curl into its original spiral shape, thereby squeezing the liquid droplets toward the evaporator (fig. 5(d)) where they can be stored in the capillary wick structure. In this system, maximum heat rejection occurs when the radiator is fully expanded. In a steady-state mode, the (d) fin spring constant could be designed so that the length of the fin would automatically adjust to balance the heat input and Figure 5.—Operation of a roll-out fin radiator. (a) Working fluid is a subcooled or saturated liquid. (b) Working fluid evaporates rejection. Roll-out fins could be made from either a thin and expands, causing fin to extend. (c) Heat radiates to space. metallic foil or plastic film with an internal spring. In the case (d) Vapor condenses, and fin curls into original shape. of metallic foil, the metal itself could be heat treated to act as a spring and provide the retraction force. Several different variations of this device have been investi- Figure 7 illustrates a concept similar to the roll-out fin; gated, including radiators that employ the previously described however in this situation, an inflatable bag system replaces the principle (Ponnappan, Beam, and Mahefkey 1984), larger multi- roll-out fin (Chittenden et al. 1988). The inflatable bags pro- component expandable radiators (Chow, Mahefkey, and posed for this concept would be made of a thin, strong, light- Yokajty 1985), and inflatable-expandable pulse power radia- weight, internally lined or coated fabric with water as the tors. Ponnappan, Beam, and Mahefkey (1984) discussed the working fluid. As illustrated, during the high heat absorption conceptual design of a 1 m long roll-out fin that could accom- phase caused by a power pulse, the radiator bags would extend modate modest peak-to-average (10:1) heat loads by varying out of the spacecraft as they expanded with vaporization of the the projected surface area. This concept has been expanded to working fluid. Then, as the spacecraft continued orbiting, the include radiator panels which utilize several of these roll-out vapor would condense as heat radiated to space. The radiator fins in parallel to form panels (fig. 6). In this application, each bags would retract during condensation, thus maintaining a of the four segments could function independently for a given constant internal saturation pressure, and they would fold into pulsed or steady heat input condition, or the four panels could the spacecraft, ready to extend again during the next power be arranged around a common vapor header and act jointly to pulse. As for the roll-out tin, this concept is characterized by a reject heat. high condensation heat transfer rate inside the radiator, low Out Vapor header

, , , R'l Roll-out fin i ' '-—J — radiator panel (fully expanded) Retraction phase Alert phase — Unexpanded B A condition

Roll-out fins joined in parallel A Closed-end fin B Open-end fin In Figure 6.—Roll-out fin expandable radiator panel concept.

operating fluid mass due to the large latent heat of vapori- zation, and high radiator effectiveness due to near isothermal operation. This type of radiator is capable of absorbing and storing substantial quantities of heat during the peak power phase of the duty cycle, and it can reject the stored heat during the longer time intervals of the cooling and retraction phase. Studies of space Strategic Defense Initiative (SDI) missions showed requirements for peak electric power in the megawatt Cooling phase Heating phase range. On the basis of these missions and their orbital cycles, Figure 7—Operational phase of high-power, inflatable bag radiator systems were required to be capable of rejecting heat radiator system. absorbed in the form of short duration pulses with peak-to- average ratios of 10 000 or more (Mahefkey 1982). Present conventional radiators are sized to reject peak heat loads, and evaporation. The vapor is collected on an expandable, variable they are turned down by lowering heat transport fluid flow to surface area (a thin metallic or plastic inner liner) on which it is reject off-peak loads. As a result, these conventional radiators allowed to condense during the time interval between pulses. are capable of near-constant-load thermal control over a range The condensate is then pumped back (or returned by other of nominally 10:1 peak-to-average heat loads. However, for means) to the coolant reservoir to be recycled. high heat load, weight-constrained applications with very high Several possible expandable containers have been proposed. peak-to-average ratios, conventional radiator designs would For high peak-heat-load pulsed radiators, low surface-to- have limited applicability (Chow and Mahefkey 1986). volume inflatable bags or bellows radiators appear promising Elliott (1984) and Koenig (1985) suggested using expand- because of their large energy storage capacity (Chow, Mahefkey, able balloon radiators to provide ultra-lightweight surfaces. and Yokajty 1985). The radiator would be constructed from a However, the utilization of expandable surfaces for cooling thin, low-density, flexible material that could be collapsed and imposes a fundamental limit on operation time. In addition, a stored in a compact form, ready for expansion during high peak severe mass penalty is associated with periodic heat release. heat loads. Because of the large volume-to-mass ratio, large Since late 1983, a collectable-expandable radiator, also known amounts of vapor could be contained during the pulse period as the expandable pulse power radiator, has been investigated and rejected through condensation and radiation during the at the Air Force Aero Propulsion Laboratory (Chow and longer interpulse period. This design results in a lightweight Mahefkey 1986). Basically, in this concept, a phase-change radiator that is very compact in the stowed position and easily material cools high power-density devices through flash protected from micrometeoroid damage, except when in use. Because of the high heat absorption and low heat-rejection rates of pulsed systems of this type, the duration of the pulse heating period trust be shorter than the interval between the pulsed, high heat load cycles. This calls for stringent restrictions on the time response characteristics. For cases requiring higher energy pulses, an expandable bellows concept has also been proposed (Chow, Mahefkey, and Yokajty 1985). The bellows concept differs from the roll-out fin and inflatable bag concepts in that a significant amount of heat energy is stored in the expanding structure. Poi cor unit Rotating Solid Radiators Solar Sensible heat capacity heat-rejection systems were first collector proposed in 1960 (Weatherston and Smith 1960). These sys- tems proposed to rotate a solid material past an internal heat source and then to space where the heat could be rejected through radiation. In a majority of these systems, heat was transferred to the solid material through either conduction or convection. An extension of this concept that has been pro- posed for high-temperature ranges such as those found in reactor cores is referred to as the radiatively cooled, inertially Figure 8.—Moving-belt radiator concept. driven nuclear generator (RING) heat-rejection system (Apley and Babb 1988). In this system, reactor waste heat is radiatively exchanger it absorbs heat, and while rotating through space transferred from a cavity heat exchanger to the rotating ring. it rejects beat by radiation. Several materials have been pro- Although at low temperatures conduction and/or convec- posed for the belt material, including homogeneous solids, tion can reduce the size of the primary/secondary interface, at or two solid belts with a phase-change material between them. higher temperatures radiative heat transfer becomes attractive. A follow-on report discusses analytical and experimental The RING power system takes advantage of the need to offset investigations of the rotational dynamics of this system along the reactor from the mission platform (for radiation field with methods for transferring heat to the moving belt, reduction) by using the space between the reactor and the deployment and stowage, and fabrication. Also, life-limiting mission platform (and the boom structural assembly) to support factors such as seal wear and micrometeoroid resistance are four counterrotating, 900 offset, coolant-carrying rings. The identified (White 1988). proposed rings are segmented, finned, thin-walled pipes that The MBR was projected to be only 10 to 30 percent as are filled with liquid lithium. massive as advanced heat pipe radiators, and it could operate The enclosed cavity heat exchanger allows a higher emis- without exposing the working fluid to space, thereby reducing sivity material to be used, and because the configuration is vaporization losses. The major technological challenge ap- protected, the primary coolant tubes can be placed closer to the pears to be maintaining the stability of a rotating belt during heat transfer surface. The cavity configuration also increases spacecraft attitude maneuvers. Although other issues, such as the hemispherical emissivity of the wall material (Siegel and long-term reliability of the roller drive mechanism, must be Howell 1980). solved, this concept compares favorably with the 5 to 8 kg/m2 space-constructable heat pipe radiator at both 300 and 1000 K. Moving-Belt Radiators A concept similar to the MBR is the liquid-belt radiator (LBR), which was also proposed by Teagan and Fitzgerald Another advanced radiator concept is the moving-belt radia- (1984). In the LBR (fig. 9), a thin screen or porous mesh tor (MBR) (Teagan and Fitzgerald 1984). This concept was structure supports a low vapor pressure liquid by capillary being developed under contract to NASA Lewis Research forces. This screen is drawn through a liquid bath where warm Center contract during the latter eighties. The basic operation liquid is picked up and retained in the screen material. The of an MBR is illustrated in figure 8, where a cylindrical belt is screen and liquid form a ribbon which is then rotated through rotated about a fixed center through some type of driving space, where heat is rejected through radiation. The advantages mechanism attached to the spacecraft. Heat collected inside the and disadvantages with this type of radiator system are similar spacecraft is transferred from a primary heat transport loop to to those of the MBR. However, in this case the liquid must have the belt through solid-to-solid conduction or directly through a very low vapor pressure (less than 10 -8 torr) over the entire convection. As the belt rotates into the spacecraft heat operating temperature range to prevent evaporative losses.

Retum pump

Film

Disk I I Nozzle

Collector trough Support structure r —Seal Rotating platform 17

Parabolic I solar dish Heat collector exchanger Figure 9.—Artist's schematic of liquid belt radiator. Figure 10.—Rotating film radiator schematic.

Several materials have been proposed, including diffusion ity, and flow regime of the film. Upon reaching the outer pump oils, gallium, lithium, and tin. The material selection circumference of the disk, the fluid is collected and returned to depends primarily on the temperature range of interest, with the the center, as illustrated. Preliminary analysis for this concept oils limited to about 350 K and the liquid metals being appli- (Prenger and Sullivan 1982) indicates that the rotating film cable over a wide temperature range, as high as 2000 K. For radiator can achieve a specific mass of 5.5 kg/kW or 3.5 kg/m2, space radiator applications, the maximum operating tempera- based on total emissivities in excess of 0.3. Hence, it cannot ture is expected to be in the 1000 K range for therm ionic power compete with advanced heat pipe radiators (to be discussed in systems. Although the proposed mode of operation is in the part II), which achieve equal or lower specific mass at surface sensible heat mode, in some situations, it may be desirable for emissivities of 0.85 to 0.9 and thus require only a third of the the LBR to operate in the latent heat mode. When this is done, surface area to reject the same amount of heat. the liquid changes phase during its traverse through space. Clearly, the mode of operation would depend on material Liquid Droplet Radiator selection, operating temperatures, and heat-rejection require- ments. Parametric analyses (Teagan and Fitzgerald 1984) indi- The liquid droplet radiator (LDR) concept retains the low- cated that the LBR could reduce the radiator mass by as much mass advantages of a disk radiator. As illustrated in figure 11, as 70 percent of space-co nstructable state-of-the-art heat pipe a warm, low vapor pressure working fluid is projected from a radiators. However, the Advanced Radiator Concepts (ARC) droplet generator, where the liquid absorbs heat, to a droplet program, to be discussed in the second part of this report, has collector, which collects the radiatively cooled droplets in a demonstrated reductions in heat pipe specific mass by a factor rotating drum (Mattick and Hertzberg 1981). Because of the of 3 to 4 over the state-of-the-art heat pipe technology used by large surface area of the droplets, this type of system has the Teagan and Fitzgerald for their basis of comparison. additional advantage of greatly reduced mass, especially with paired modules, which eliminates the need for a long return Rotating Film Radiator loop for the liquid. These paired modules would be connected by a structural tie rod, thus maintaining the proper alignment Figure 10 depicts another advanced radiator concept that between droplet generators and collectors. The generator is a uses a thin liquid film: the rotating film radiator (Song and pressurized plenum with an array of holes or nozzles to form Louis 1988). As shown, this concept uses a rotating disk with liquid jets that break up into droplets via surface tension a thin film of liquid flowing radially. Initially, the proposed instability. A piezoelectric vibrator, which rapidly varies the working fluid, toluene, is injected at the center and the flow is pressure of the fluid, can also be employed to control the drop split equally between the two surfaces of the disk. The fluid then size and spacing. Droplets could be generated and collected, spreads into thin films which facilitate radiation of heat to and heat transferred to the liquid, with only modest extensions space. The disk's rotational speed controls the thickness, veloc- of conventional technology. LDR's have a large surface area Droplet

Figure 11.—Dual-module solar power satellite with liquid droplet radiator. unit per unit mass, or low mass per unit radiating area, but systems that have near constant heat-rejection temperatures this advantage is offset to some extent by the lower effective (Juhasz and Chubb 1991). However, they are not compatible emissivity of the droplet sheet than that of advanced heat pipe with closed-cycle gas turbine power systems, which must reject radiators. However, proponents during the last decades argued heat over a broad temperature range (Juhasz and Chubb 1991; that with low vapor pressure liquids, which are available over Juhasz, El-Genk, and Harper 1993). A recent status report on a wide radiating temperature range (250 to 1000 K) with LSR development (Chubb, Calfo, and McMaster 1993) sum- negligible evaporation loss (silicone oils, 250 to 350 K; liquid marizes the work done on sheet stability and points out the need metal eutectics, 370 to 650 K; and liquid tin, 550 to 1000 K), the to conduct sheet emissivity measurements and to develop a LDR could be adapted for a wide range of heat-rejection sheet fluid collector before a viable LSR can be demonstrated. applications (Elliott 1984). A governing factor in the design of LDR's is the mass loss via Curie Point Radiator evaporation. The mass required to replenish the evaporation must be included in the overall radiator mass for comparison The Curie Point Radiator (CPR) maintains the low mass with other systems. However, for rejection temperatures advantage of the LDR and is similar in operation with one major between 300 and 1000 K, liquids are available with low exception. The CPR uses a large number of small, solid ferro- enough vapor pressures that evaporation losses can be consid- magnetic particles (Carelli et al. 1986). These particles are erably smaller than the radiator mass, even for operational heated to a temperature above the Curie point, the point at lifetimes of 30 years. Thus droplet radiators were considered which a ferromagnetic material loses its magnetic properties, suitable for a wide range of applications—from heat rejection and are ejected from the heat source toward a magnetic field. As in high-temperature thermal engines, where rejection tempera- the particles radiate heat to space and cool, they regain their tures might be in the 500 to 1000 K range, to cooling of magnetic properties and are collected by a magnetic field photovoltaic cells and heat rejection from refrigerators, where collector. The CPR has all of the advantages of the LDR, such rejection temperatures would be in the 250 to 350 K range as low mass-to-radiating-area ratio, reduced mass for microme- (Mattick and Hertzberg 1981). teoroid protection, and a small mass of radiating particles that An extension of the LDR, the liquid sheet radiator (LSR), has represents only a minor fraction of the total. In addition, the also been proposed. The operation of this type of system is unique characteristics of the ferromagnetic particles result in similar to that of the LDR with the exception that a continuous several other significant advantages, including (1) a particle liquid sheet, rather than a multitude of individual droplets, is inventory that can be actively controlled, thereby reducing the used to reject heat. Because the narrow slits that produce sheet loss of particles, (2) particles that can be coated to increase flow can be fabricated without the precision machining tech- surface emissivity (a value as high as 0.9 can be achieved with niques required for a large number of small orifices, this system SiC coating), and (3) elimination of the need for strict tempera- reduces the level of technology development required. In ture control. The key disadvantage is the possibility of mag- addition, the LSR requires less pumping power because of the netic perturbations to other components of the spacecraft. Also, reduced viscous losses, and it offers a simplified collection the problems of transferring spacecraft reject heat to the particle system because of a self-focusing feature (Chubb and White stream and the actual mass transport of the particles through the 1987). Both the LDR and the LSR are compatible with power power system heat exchanger have not been solved. Rotating Bubble-Membrane Radiator To operate effectively in space, the rotating bubble-mem- brane radiator will include design features to minimize damage Perhaps the most promising alternative to space-cons tructable and to mitigate any coolant losses that may result from meteor- heat pipe radiators, after LDR technologies, is the rotating oid and space debris impact. New high-strength, low-weight bubble-membrane radiator (Webb and Antoniak 1988). A fiber and metallic alloy cloths show excellent promise for rotating bubble-membrane radiator functions as a two-phase, inhibiting micrometeoroid penetration of the rotating bubble- direct-contact heat exchanger. This hybrid radiator design membrane radiator (Webb and Antoniak 1988). In addition, incorporates the high surface heat fluxes and isothermal oper- design options are being considered that would seal membrane ating characteristics of conventional heat pipes along with the penetrations and reduce coolant losses from micrometeoroid low system masses normally associated with LDR's. As de- penetrations. Selection of materials for the thin-film membrane picted in figure 12, a two-phase working fluid enters the bubble will be dictated by the desired operating temperature. Candi- through a central rotation shaft where it is sprayed radially from date materials include carbon-epoxy compounds, silica, a central nozzle. This combination of liquid droplets and vapor alumino-borosilicate, or silicon carbide cloth with metallic moves from the central portion of the bubble toward the inner liners (Sawko 1983), and niobium-tungsten composites. The surface, transferring heat by both convection and radiation. As final selection of the envelope material will depend on the the droplets move outward, they increase in size because vapor radiator fluid and its intended operating temperature. Pump condenses on the droplet surface and droplets collide with each selection also will be determined by the working fluid. Electro- other. Upon striking the inner surface of the radiator, the magnetic pumps are possible candidates for liquid metal cool- droplets form a thin surface film. This film then flows toward ants, and mechanical or electric pumps are favored for other the equator because of the rotationally induced artificial grav- applications. ity. Heat transfer between the fluid and bubble radiator then becomes a combination of conduction and convection. As the Technology Comparisons fluid reaches the equator of the sphere, it is collected in a gravity well and pumped back to repeat the process. Among the various heat pipe technologies considered for advanced heat-rejection systems (see table I), external artery Return pump and conventional, axially grooved heat pipes have the greatest collection Radiating heat-rejection capabilities. For high peak loads, expandable

Return surface roll-out fin radiators with pulsed heat absorption capability piping/structural offer a considerable weight savings over conventional tube and support , fin radiators. However, further study is necessary to reduce their vulnerability to micrometeoroids and to improve the header/fin heat exchanger design and the operating character- istics (Ponnappan, Beam, and Mahefkey 1984). Among the inflatable radiator concepts previously developed, a retractable bellows configuration with a stationary sponge appears to be the best candidate. Although previous analyses indicate that this type of system has good dynamic stability and excellent thermal behavior, additional investigations are required. Central Several other advanced radiator systems have been proposed rotation to reduce the mass requirement. Among these, the most actively shaft — pursued through the late eighties was the LDR, where a large number of submillimeter liquid droplets constitute the radiat- ing surface. Although the LDR has the potential for substantial reduction in mass versus conventional radiator systems, prob- spray nozzle lems associated with inventory losses due to vaporization, aiming inaccuracies, and splashing on the collector—which tend to increase the total system mass—must be addressed in future work. Another disadvantage of the LDR that must be resolved is that it is not suitable for missions requiring high maneuverability during full-power operation. Similarobserva- tions apply to the LSR, although it would be much easier and cheaper to fabricate the coolant fluid injectors, for this concept. The MBR concept compares favorably with heat pipe radiators Figure 12.—Boom-mounted rotating bubble-membrane radiator. on the basis of specific mass. The biggest advantage is achieved

9 TABLE I. - COMPARISON OF ADVANCED RADIATOR CONCEPTS

Criterion' SCR ROF RSR MBR LBR RFR LDR/LSR CPR RBMR

Weight Mod. Mod. High Mod. Mod. Mod. Low Low Low

Reliability Mod. Avg Mod. Mod. Mod. Good Mod. Excel. Good

Maintenance required Low Mod. Low Mod. Mod. Low Mod. Mod. High

Technology readiness Excel. Mod. Good Poor Poor Mod. Mod. Poor Poor

Life expectancy Good Good Mod. Mod. Poor Mod. Excel. Unkn. Mod.

System complexity Low High Mod. High Mod. Mod. High High High

Area required High High Mod. Mod. Mod. Mod. Mod. Mod. Low

Performance Excel. Unkn. Good Unkn. Unkn. Mod. Good Unkn. Mod.

Life cycle cost Low Unkn. Unkn. Unkn. Unkn. Unkn. Unkn. Unkn. Low

Micrometeoroid Mod. High Low Mod. Mod. Mod. Low Low High vulnerability

'SCR space-constructible radiator ROF roll-out 5n radiator RSR rotating solid radiator MBR moving belt radiator LBR liquid belt radiator RFR rotating film radiator LDR liquid droplet radiator CPR Curie point radiator LSR liquid sheet radiator RBMR rotating bubble membrance radiator with the hybrid belt system that exploits the phase-change collector with a magnetic field generator and an effective heat potential of an LBR, yet offers high surface emissivities over a exchanger to transfer heat to the solid particles from the broad range of temperatures. However, control of the belt shape working fluid. under microgravity operating conditions and long-term reli- ability of the roller drive system are serious problems. Accord- ing to its proponents, the CPR, which utilizes the ferromagnetic Part II.—Highlights of the NASA Lewis properties of radiating particles, represents a unique concept Civil Space Technology Initiative Thermal that offers significant advantages in mass, high reliability, and a practically unlimited temperature range (Carelli et al. 1986). Management Program However, some key drawbacks, including heat transfer to and actual mass transport of the particles through the power system Civil Space Technology Initiative (CSTI) thermal manage- heat exchanger, and the possibility of magnetic perturbations to ment related work at Lewis (to be concluded during 1994) has other spacecraft components, had not been resolved at the time been an integral part of the NASA CSTI High Capacity Power of program termination. Program and, specifically, of the Tri-Agency (Department of Although the SCR is the most developed of the concepts and Defense, Department of Energy, and NASA) SP-100 nuclear is a proven, reliable technology, additional investigations into reactor space power program (Winter 1991). the behavior during heat pipe startup from a frozen working The goal of the Lewis thermal management effort (Juhasz fluid state and the effect of on-orbit accelerations are needed. 1991) is to develop near-term space radiator and heat-rejection Although rotating bubble-membrane radiators and rotating system concepts, optimized for a spectrum of space power film radiators presently lack the technical maturity of heat pipe conversion systems for planetary surface (lunar base) and radiators, rotating machinery and shaft vapor seals to space nuclear propulsion applications for deep-space, long-duration have proven effective and do not require additional develop- missions needed for the Space Exploration Initiative (Bennett ment. With MBR's, seals must wipe off the working fluid and Cull 1991). The power, or energy, conversion system without allowing leakage or damage to the belt as the belt concepts range from static systems, such as thermoelectric or exits the heat exchanger. In contrast, the LDR, CPR, and thermionic, to dynamic conversion systems based on heat rotating bubble-membrane radiator require presently nonexist- engines, such as the Stirling engine or the closed-cycle gas ent technologies that must be completely developed, tested, and turbine, also known as the Closed Brayton Cycle (CBC). qualified. LDR's require a droplet generator/collector combi- Although the principal heat sources for these systems are nation with a high degree of aiming accuracy. CPR's require a nuclear (Juhasz and Jones 1987), the technology being

10 developed for the heat-rejection subsystem is also applicable to puter code development for predicting heat pipe performance, low-Earth-orbit (LEO) based dynamic power systems with both under steady state and transient operating conditions, solar energy input, using a concentrator and heat receiver. along with experimental testing to validate the analytical pre- Brandhorst, Juhasz, and Jones (1986) studied such systems as dictions. alternatives to photovoltaic power systems. The performance Continued research on radiator surface treatment techniques goals for the advanced radiator concepts being developed are (surface morphology alteration) is contributing to the in-house lower radiator mass (specific mass of 5 kg/m 2 or lower) at a advanced development program. This research is aimed at surface emissivity of at least 0.85 over the entire operating enhancing surface emissivity and resistance to atomic oxygen temperature range, greater survivability in a micrometeoroid or attack. space debris environment (up to 10 years), and a subsystem The development of new radiator materials with high strength- reliability of 0.99 or higher. These performance goals may be to-weight ratios and high thermal conductivity (such as carbon- realized by radiator segmentation and parallel redundancy, carbon composites for lightweight radiator fins) is another using a large number of heat pipes. Achieving these goals may major objective. Figure 14 shows the project plan. Note that reduce the SP-100 radiator specific mass by a factor of 2 or because of funding constraints, the development of the far-term more over the original baseline design and may lead to even innovative radiator concepts discussed in Part I, such as the greater mass reductions for radiators used in contemporary LDR and the MBR, is not being actively pursued. Instead, spacecraft. technologies that could be developed before the end of the The project elements (fig. 13) include development of decade are being concentrated on for both surface power and advanced radiator concepts under Lewis-managed contracts nuclear electric propulsion (NEP) applications. Near-term and a NASA/Department of Energy interagency agreement, as applications to small spacecraft and technology transfer to well as in-house work directed at radiator design for optimum possible terrestrial uses are also being considered. power system matching and integration. In-house and The remainder of this report reviews the major project university-supported heat pipe research and development elements, concentrating on the contracted efforts that account also is being carried out. This work includes analytical com- for the major portion of the baseline budget.

Heat pipe Advanced Radiator Concepts contracts • Analysis codes • Testing

• Phase I, II • University of New Mexico, Space Power Inc. (SPI), University of California, Rockwell International, (Los Angeles), Hughes, GE Wright State University, Goal NASA Lewis • Phase III, IV < 5 kg/m 2, e >_ 0.85 SPI, Rockwell International • Los Alamos National Lab., > 0.99 rel NASA Lewis, Wright Research 10 r life • Ultra-light weight fabric y and Development Center, heat pipe development Phillips Laboratory/Space Department of Energy/ Thermal and Power Technology Pacific Northwest Laboratory

Surface morphology Composite materials • Emissivity >0.85 • Arc texturing •Refractory + Gr/Cu • LDEF input • Carbon/carbon Radiator design Lewis (Electro-Physics Branch) & integration • Lewis (Materials Division) • SPI, Rockwell International

System analysis—Lewis (Power Systems Integration Office) Figure 13.—Lewis thermal management project elements. (LDEF, the Long-Duration Exposure Facility, was launched in 1984 and retrieved from space in 1990.) Feasibility Demonstrations Phase IV contracts To SP-100 project Advanced Radiator Concepts contractors _ Phase III 875 K concept demo ARC (Space Power Inc. (SPI), complete 600 K concept demo Rockwell International) (Components) To SP-100 TEM pump

Lewis High-conductivity composite fin development fabrication and test (NASA Lewis, Applied Sciences Inc., Science Applications International Co.)

Fabric-foil/heat pipe P. H2O SiC fabric Lewis: (Department of Energy/Pacific heat pipe 3-mil (0.075 mm) vacu Northwest Laboratory) Cu foil test

89 90 91 92 93 94

Figure 14.—Thermal management project plan.

Advanced Radiator Concepts Development Contracts version systems, and a low-temperature option (500 to 600 K) applicable to Stirling power conversion systems. Rockwell has The Advanced Radiator Concepts (ARC) contractual devel- focused on heat-rejection technology for the higher tempera- opment effort is aimed at the development of improved, light- ture thermoelectric power systems. weight space heat-rejection systems, with special emphasis on space radiator hardware, for several power system options, Contract NAS 3-25208 With SPI including the thermoelectric and Free Piston Stirling (FPS). The targeted improvements will lead to lower specific mass Among the advanced concepts proposed by SPI are the (mass per unit area) at high surface emissivity, higher reliabil- Telescoping Radiator (Begg and Engdahl 1989 and Koester ity, and higher survivability in a natural space environment, and Juhasz 1991) for multimegawatt thermoelectric or liquid thereby leading to longer life for the power system as a whole. metal Rankine power systems (fig. 15), and the Folding Panel As stated earlier, specific objectives are to reduce specific mass Radiator (Koester and Juhasz 1991) (fig. 16) for the 500 to to <5 kg/m2 with radiator surface emissivities of 0.85 or higher, 600 K heat-rejection temperature range. The latter concept was at typical radiator operating temperatures and with reliability based on a pumped binary lithium/sodium potassium (Li/NaK) values of at least 0.99 for the heat-rejection subsystem over a loop, motivated by a desire to avoid the need for mercury heat 10-year life. Although the above specific mass figure does not pipes (HgHP) and their potential adverse effects on spacecraft include the pumps and the heat transport duct, it represents electrical systems. Such HgHP radiators were originally planned about a factor of 2 mass reduction over the baseline SP-100 for an FPS power system that rejects heat in the above tempera- radiator, and it implies an even greater mass savings for the ture range. state-of-the-art heat-rejection systems used in current space- craft applications. Li/NaK Binary Pumped Loop Phases I, II, and III of the ARC contracts have been com- pleted by both contractors: Space Power Inc. (SPI) of San Jose, A detail of a typical Li/NaK heat-rejection loop, which also California, and Rockwell International of Canoga Park, Cali- uses high-conductivity fins for heat transport is illustrated in fornia. On the basis of the phase III results, both contractors figure 17. As indicated in figure 18(a), the advantage of using were selected to proceed into phase IV—component-level a Li/NaK mixture, rather than NaK alone (melting point, development, fabrication, and demonstration—to be accom- 261 K), lies in its combining the high heat capacity and low plished over a 2-year period and to be concluded by early 1994. pumping power of Li (melting point, 452 K) with the liquid SPI is developing both a high-temperature heat-rejection pumping capability of NaK, down to its freezing temperature of option (800 to 830 K) applicable to thermoelectric power con- 261 K.

12 Reactor

Figure 15.—Multimegawatt, telescoping cylinder heat pipe radiator concept; longitudinal potassium heat pipes with circumferential heat pipe fins.

Figure 16.—Folding-panel heat pipe radiator concept.

13 Insulated

Electromagnetic ., i pump

Graphite/carbon conduction fin Heat pipe

Li

Li/NaK —NaK

Normal operation Partial freezing/thawing Totally frozen condition at startup/restart Figure 17.—Pumped Li/NaK binary loop radiator concept.

To illustrate the operation of this binary loop during system semipermeable membranes to NaK under certain operating startup and shutdown, a brief explanation is in order. During conditions. startup (with Li frozen), liquid NaK would be pumped through A two-dimensional computer analysis of the cooldown the inner cores of radiator tube passages in hydraulic contact (freeze) and warmup (thaw) processes also has been completed, with the frozen layers of Li coating the inner passage surfaces. including color graphics output. Although the flat flow channel As the NaK is heated during power system startup, it eventually cross-sectional geometry assumed in the analysis deviated melts the solid Li annuli by direct-contact, forced-convection from the cylindrical flow channel used in the experimental heat transfer. The melted Li progressively mixes with the NaK loop, this computer code nevertheless permitted visualization to form the all-liquid Li/NaK coolant. Conversely, on shut- of the basic phenomena within the binary loop during the down of the power system, the molten Li with its higher warmup and cooldown periods. A video tape of the analysis freezing point will selectively "cold trap," or freeze, on the illustrates several cases with and without plugging of the flow inner passage surfaces as their temperatures drop below the channel due to Li freezing over the entire channel cross section. 452 K freezing point, while the NaK continues to be pumped in Time and funding permitting, freeze-thaw behavior at higher its liquid state through the inner cores of the radiator passages. than 50 vol % Li will also be briefly explored (Koester and Tests conducted thus far, using a trunnion-mounted test loop, Juhasz 1994). which can be rotated about pitch and roll axes to isolate gravity effects (fig. 18(b)), have demonstrated the feasibility of the High-Conductivity Fin Development concept up to Li volume fractions of 50 percent. In particular, the feasibility of a heat-rejection system based on a binary Li/ Progress also was made by SPI in its work with innovative NaK pumped loop was demonstrated during transient operat- subcontractors, namely Applied Sciences Inc. (ASI) and Sci- ing conditions, representative of both the cooldown (Li freez- ence Applications International Co. (SAIC), who have demon- ing) and warmup (Li melting) phases of typical alkali metal strated considerable success in the development and fabrication heat-rejection pumped loops. At certain operating conditions, of high thermal conductivity composite materials for space the thawing process had to be controlled very closely to avoid radiator fin applications. In particular, ASI has produced a plugging the test section flow passage downstream of the melt composite by chemical vapor deposition and densification of front at certain operating conditions. Current efforts focus on closely packed (up to 60 vol %) vapor-grown carbon fibers widening the operating envelope by a variety of techniques. which was shown to have high thermal conductivity (near 560 One of these involves the use of fine mesh screens which act as Whn•K) at room temperature, and a density of 1.65 g/cm3.

14 Ar Vacuum Level Ar detector Vacuum

Ar 2000 .8 8 Vacuum Cooling water 1750 .7 7 Li charge Heat sink 1500 .6 6 tank 3 U Ar Density Ar Vacuu m 'a) 1250 .5 '' 5 Electro- Y vacuum Volume Argon a magnetic a 1000 •4 4 pump compen- reservoir Q a Heat NaK sator capacity ,3 3 charge E 750 p m CL a^ tank T*Water 2 drain 500 .2 Magnetic tank Pumping flowmeter power—,, 250 .1 1 Ar Vacuum 0 0 0 20 40 60 80 100 Drain NaK content in Li, % tank (a) (b)

Figure 18.—Binary Li/NaK radiator development. (a) Pumping characteristics of Li/NaK mixtures. 5-MW transported; fluid AT = 100 °C; 15 parallel loops with 5-cm diameters. (b) Test loop.

Similarly, SAIC has successfully fabricated composites using figure 19. Because each of the "petals," or radiator panels, is short (about 0.01 m long) vapor-grown carbon fibers with a composed of a large number (384) of variable length carbon- density of 1.6 g/cm3 and a demonstrated conductivity of carbon (C–C) heat pipes mounted transverse to the panel axis, 470 W/m•K at operating temperatures near 600 K. A compari- a major objective of this effort is to develop these integrally son of these properties with those of copper (density, 8.9 g/cm3, woven graphite/carbon tubes with an internal metallic barrier and thermal conductivity, 380 W/m •K) reveals a significantly that is compatible with the intended potassium working fluid. higher conductivity-to-density ratio for these composites. C–C heat pipe tube sections with integrally woven fins were Use of composite materials with specific thermal conductiv- fabricated under phase III (Rovang et al. 1990). Highlights of ity values at these levels for heat pipe fin applications permits the fabrication process are illustrated in figure 20. Because of increasing the fin length at constant fin efficiency, and it its low cost, commercial availability, and ease of weaving, a therefore has the potential of reducing radiator specific mass by T-300 fiber was selected for this demonstration of C–C heat over 60 percent for radiators that are radiative heat transfer pipe preform fabrication. This polyacrylonitrile (PAN) fiber surface limited. Recently, technology for joining the high- was judged to represent a tradeoff between high elastic modu- conductivity fins to the heat pipes by advanced brazing or lus, and consequently ease of handling and weaving, and welding techniques and processes that will lead to even higher medium thermal conductivity (80 W/m•K), in contrast to some composite thermal conductivity values also have been demon- very high conductivity fibers which, however, may be brittle strated (Denham et al. 1994). and difficult to weave. Several fiber architectures were investigated before settling Contract NAS 3-25209 With Rockwell International on an angle interlock, integrally woven concept. In this design, the axial fiber bundles, referred to as warp weavers, are woven A sketch of the Petal-Cone radiator concept being developed in an angle interlock pattern, repeatedly traversing from the at Rockwell International (RI)(Rovang 1988) is shown in inner diameter to the outer diameter surface of the tube. An

15 C_A ---

-eployed position envelope for stowed radiator

C11U ,. Iy (a) (b) Figure 19.—Petal-cone heat pipe radiator concept. (a) Heat pipe. (b) Cone radiator (12 panels).

Interlocking yarns

3D;

Continuous pipe preform weavingL2^ Multiple preform impregnation Batch nesting ' partial view

Top view graphitization furnace Multiple cavity platen molding Final machining Carbon -carbon S^ cleanup and trim densification processing i.d. surface coating

Figure 20.—Integral, pinned graphite-carbon heat pipe fabrication process. unfilled "Novo Iack"/resole prepreg resin was selected for tal improvements over each previous coating attempt, but a prepregging the woven preforms followed by a low-pressure, constant thickness coating over the full-length of the tube high-temperature impregnation and carbonization process for without any flaws or imperfections could not be achieved. densification of the composite. Because of these coating problems, RI was directed at the Considerable progress was made in the development of beginning of phase IV to use a thin-walled metallic liner to internal metallic coatings to ensure compatibility of the heat safely contain the heat pipe working fluid instead of using the pipe surface with the potassium working fluid. A coating metallic coating that had been under development during phase consisting of a 2 to 3 ).un rhenium sublayer with a 70 to 80 pm III. Another advantage of the metal liner approach was that the niobium overlayer emerged as the final recommended coating heat pipe evaporator could be formed by simple extension of design. Because of funding and time limitations, however, this the liner beyond the C–C shell. final coating design and the recommended method to achieve Concurrently with this task, a high-temperature braze or it, a novel chemical vapordeposition process that uses a moving other joining process was developed to ensure good mechanical heat source, could not be fully implemented during phase III. and thermal contact between the thin metallic liner and the Other coating approaches which were tried achieved incremen- C- C internal tube surface. Bonding of the entire liner surface to

16 a finned 0.3 m long C—C tube also has been achieved by use of pipe fabrication discussed above, and it is expected to have ternary braze alloys, such as Silver ABA or Cusil ABA. This broad application beyond the scope of this program. The water was necessary to prevent partial separation and local collapse heat pipes fabricated for NASA Lewis by Pacific Northwest of the liner at conditions prior to launch, where the external Laboratories were evaluated for performance and reliability at atmospheric pressure exceeds the internal pressure of the demanding operating conditions, including operating pressures working fluid. up to 25 bar. Tests were conducted with and without wicks, Concerning the integral fin weaving process, using T-300 with the heat pipes in various gravity tilt orientations from fibers, significant improvements were made, which lead to the vertical to horizontal. In addition, several wick designs were elimination of a troublesome internal cusp formed at the fin- tested forcapillary pumping capability, both in ground tests and tube interface. This was accomplished by changing the weave in low-gravity, KC-135 aircraft testing (Antoniak et al. 1991). architecture so that the outer, rather than the inner, plies were Future work in this area needs to focus on perfecting the heat used to form the fins. pipe fabrication procedure by using very thin (I to 2 mil (0.025 In addition to the fabrication and testing of a complete 2.5 cm to 0.05 mm)) foil liners, internally texturized by exposure to o.d. heat pipe with a niobium/zirconium alloy liner and a T-300 high pressures. With sufficient development, the textured composite shell, the fabrication of a higher conductivity com- internal surface may provide the required capillary pumping posite (P95WG) finned heat pipe shell was also completed. between the condenser and evaporator. Because of the high With a measured conductivity of over 300 W/m •K for this operating pressures required with water (over 16 atm), hyper- composite, fin length could be doubled from 2.5 to 5 cm, thus velocity and ballistic velocity simulated micrometeoroid reducing the specific mass (for one-sided heat rejection) to impact tests on thin-walled pressurized tubes enclosed in wo- 2.9 kg/m2, in comparison to 4.2 kg/m2 for the T-300 com- ven fabric will be needed to ascertain if secondary fragments posite. For two-sided heat rejection, as occurs with flat plate from a penetrated heat pipe could cause neighboring heat pipes radiators, the above specific mass values are effectively cut in to fail. Another major challenge will be to design and fabricate half (Juhasz and Bloomfield 1994). A recent update summariz- a heat pipe with high-conductivity, lightweight fins as a first ing the fabrication and testing of the Rockwell/NASA C—C step toward lightweight radiator panels. heat pipe is given by Rovang, Hunt, and Juhasz (1994). Supporting Project Elements Lightweight Advanced Ceramic Fiber Heat Pipe Radiators Space limitations prevent a detailed discussion of the The objective of this joint NASA Lewis, Air Force (Phillips remaining project elements referred to in figure 13. However, Laboratory, Kirtland Air Force Base), and Department of a brief paragraph highlighting these activities is warranted. Energy (Pacific Northwest Laboratories) program was to dem- As mentioned previously, the system integration studies per- onstrate the feasibility of lightweight, flexible ceramic fabric formed at Lewis guide the overall thermal management work (such as aluminum borosilicate) metal-foil-lined heat pipes for by providing the proper application for it. As shown by Juhasz a wide range of operating temperatures and working fluids. and Chubb (1991), for example, an LSR with lighter specific Specifically, the Lewis objectives were to develop this concept rnass than a heat pipe radiator will not necessarily benefit all for application to Stirling space radiators with operating tem- power conversion systems equally. As discussed in the refer- peratures below 500 K, using water as the working fluid. The ence, the LSR (or the LDR) concept is not suited to the specific mass goals for these heat pipes were <3 kg/m 2 at a relatively steep heat-rejection temperature profile of a Closed- surface emissivity of at least 0.80. Brayton-Cycle power system. However, it does work well with Several heat pipes were built with titanium or copper foil a Stirling power system, which rejects heat at a near constant material for containment of the water working fluid. A heat pipe temperature. For a further example of how radiator-power with an 8 mil (0.2 mm) titanium liner, designed to operate at system integration studies are used to ascertain radiator- temperatures up to 475 K, was demonstrated at the 8th Sympo- induced power system performance degradation, refer to sium on Space Nuclear Power Systems (Antoniak et al. 1991). figure 21. The curves illustrate the reduction in power output An innovative Uniskan Roller Extrusion process was devel- and efficiency for both Brayton and Stirling power systems, oped at Pacific Northwest Laboratories and used to draw 30 mil resulting from a reduction in cycle temperature ratio due to a (0.75 mm) wall tubing to a 2 mil (0.05 mm) foil liner in one pass. loss of radiator area. Such area loss may be caused, forexample, Moreover, this process eliminated the need for joining the thin by micrometeoroid damage. It should be noted that even with foil section to a heavier tube section for the heat pipe evapora- a loss of 50 percent of radiator area, a Stirling engine can still tor, which needs to be in tight mechanical and thermal contact produce over 75 percent of its design power, whereas a Closed- with the heat-rejection system transport duct. End caps are Brayton-Cycle system produces over 65 percent of its rated attached to the evaporator and condenser ends of the tubular output. liner by specialized brazing or welding techniques. The liner A typical example of radiator surface morphology altera- fabrication technique also was applied to the Rockwell heat tion by arc texturing for emissivity enhancement purposes is

17

1.0

.9

U C N Cv 8 N ZF NC W N N m .7

d 3 O a 3.6 O C a ^ d 5

.4 1 1 1 1 1 1 1 1 1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 Fraction of reference radiator area Figure 21.—Sample results from radiator-power system integration study.

60-Hz variableEmittance at 1159 K Water cooling power supplyly Emittance at 322 K and insulation, '4 1.00

.80

.60 ^^^ c m Carbon Variable frequency and electrode variable waveform w .40 power supply Sample .20

Vacuum based 0 sample holder 0 10 15 20 and electrode Texturing current, A (a) (b) Figure 22.—Surface emittance enhancement by arc texturing. (a) Texturing equipment setup. (b) Emittance of graphite/copper samples.

18 shown in figure 22. Along with the arc-texturing apparatus in figure 22(a), the adjoining bar chart (fig. 22(b)) shows the Acknowledgments surface emittance achieved with various arc current values for graphite-copper samples produced under the in-house materi- als program. Additional details of the emittance enhancement The authors gratefully acknowledge the support provided by process and measurement techniques are discussed by Rutledge, the Texas A&M University Center for the Commercialization Forkapa, and Cooper (1991). Rutledge, Hotes, and Paulsen of Space Power, the U.S. Naval Research Laboratory, Wash- (1989) detail the emittance enhancement of C–C composite ington, D.C., and the NASA Lewis Research Center, Cleve- surfaces by atomic oxygen beam texturing. A brief overview land, Ohio. The dedicated work and excellent cooperation of of the same topic is included in a CSTI status report the ARC contract participants—Space Power Inc., Rockwell (Winter 1991). International, Pacific Northwest Laboratories, Science Appli- Heat pipe performance modeling, both under steady cations International Co., and Applied Sciences Inc.—contrib- state and transient operating conditions, is being concluded uted greatly to the success of the Advanced Radiator Concepts at Lewis and under university grants with the University of program. California, Los Angeles; the University of New Mexico; and Wright State University. The objective of this work is to develop a capability to analytically predict transient operation of heat pipes, particularly during startup and cooldown. An References especially important feature of this work is the development of Alario, J. 1984. Monogroove Heat Pipe Radiator Shuttle Flight Experiment: an analysis code that can model startup for a variety of working Design, Analysis and Testing. SAE Paper 840950. fluids (including water and liquid metals) when the working fluid is initially frozen. Working versions of the codes have Antoniak, Z.1., Webb, B.J., Bates, J.M., Cooper, M.F., and Pauley, K.A. 1991. been developed, and validation of predicted performance by Testing of Advanced Ceramic Fabric Wick Structures and Their Use in a Water Heat Pipe. Presented at the National Heat Transfer Conference, laboratory testing of heat pipes is under way at Lewis, Minneapolis, MN, July 28-31. Los Alamos National Laboratory, and Wright State University. Efforts also have been initiated to compile an experimental Apley, W.J., and Babb A.L. 1988. Rotating Solid Radiative Coolant System for database by a systematic literature search and by close commu- Space Nuclear Reactors. Technical Report PNL–SA-15433, Pacific nication with other researchers in the field. Northwest Laboratories, Richland, WA. (Also, AIAA Paper 88-3189). Begg, L.L., and Engdahl, E.H. 1989. Advanced Radiator Concepts—Phase 11 Final Report for Spacecraft Radiators Rejecting Heat at 875 K and 600 K. Concluding Remarks NASA CR-182172.

The NASA Lewis Research Center's Civil Space Tech- Bennett, G., and Cull, R.C. 1991. Enabling the Space Exploration Initiative: NASA's Exploration Technology Program in Space Power. AIAA Paper nology Initiative (CSTI) Thermal Management Program was 91-3463. designed to combine a number of project-oriented elements in order to accomplish the overall objective of reducing radiator Brandhorst, H.W., Juhasz, A.1., and Jones, B.I. 1986. Alternative Power specific mass by at least a factor of 2 at a subsystem reliability Generation Concepts in Space. NASA TM-88876. of 0.99 over a 10-year service life. Although the main focus was Carelli, M.D., Gibson, G., Flaherty, R., Wright, R.F., Markley, R.A., and on support of the SP-100 program by advances in heat- Schmidt, J.E. 1986. A Novel Heat Rejection System for Space Applica- rejection technology, the concepts and hardware developed tions: The Curie Point Radiator. Proceedings of the 21st Intersociety under this program are expected to benefit space power systems Energy Conversion Engineering Conference: IECEC'86, ACS, Wash- in general, ranging from solar photovoltaic or solar dyna- ington, D.C. Vol. 3, pp. 1875-1880. mic systems, with a power level of a few kilowatts to future multimegawatt power systems with nuclear heat sources for Chittenden, D., Grossman, G., Rossel, E., Van Etten, P., and Williams, G. 1988. High Power Inflatable Radiator for Thermal Rejection From Space planetary surface and nuclear (electric) propulsion applica- Power Systems. Proceedings of the 23rd Intersociety Energy Conversion tions. In spite of the termination of the SP-100 program, its Engineering Conference 1988 IECEC: ASME, New York, Vol. I, major subsystem technology advances, especially in the ther- pp. 353-358. mal management area, are judged to be ready for on-orbit Low Temperature demonstration before the end of the decade. Thus, NASA's Chow, L.C., Mahefkey, E.T., and Yokajty, J.E. 1985. Expandable Megawatt Pulse Power Radiator. AIAA Paper 85-1078. and the nation's long-term goals in space exploration and utilization may be realized sometime during the next century. Chow, L.C., and Mahefkey, E.T. 1986. Fluid Recirculation, Deployment and In the mean time, terrestrial and small spacecraft applications Retraction of an Expandable Pulse Power Radiator for Spacecraft Power of these technologies also will be pursued. Supplies. AIAA Paper 86-1322.

19 Chubb, D.L., and White, K.A. 1987. Liquid Sheet Radiator. NASA TM-89841 Koester, J.K., and Juhasz, A.J. 1994. NASA Advanced Radiator Technology (Also, AIAA Paper 87-1525.) Development. Presented at the 1 1 th Symposium on Space Nuclear Power and Propulsion, Albuquerque, NM, Jan. 9-13. Chubb, D.L., Calfo, F.D., and McMaster, M.F. 1993. Current Status of Liquid Sheet Radiator Research. NASA TM-105764. Kossan, R., Brown, R., and Ungar, E. 1990. Space Station Heat Pipe Advanced Radiator Element (SHARE) Flight Test Results and Analysis. AIAA Denham, H.B., Koester, K.J., Clarke W., and Juhasz A.J. 1994. NASA Paper 90-0059. Advanced Radiator C-C Fin Development. Presented at the I 1 th Sympo- sium on Space Nuclear Power and Propulsion, Albuquerque, NM, Leach, J.W., and Cox, R.L. 1977. Flexible Deployable-Retractable Space Jan 10-14. Radiators. AIAA Paper 77-764.

Edelstein, F. 1987. Thermal Bus Subsystem for Large Space Platforms. Mahefkey, T. 1982. Military Spacecraft Thermal Management, The Evolving Technical Report TR-3147-02, Grumman Aerospace Corporation, Requirements and Challenges. AIAA Paper No. 82-0827. Bethpage, N.Y. Mattick, A.T., and Hertzberg, A. 1981. Liquid Droplet Radiators for Heat Elliott, D.G. 1984. Rotary Radiators for Reduced Space Powerplant Temper- Rejection in Space. Journal of Energy, vol. 5, no. 6, pp. 387-393. (Also, atures. Proceedings of the 1st Symposium on Space Nuclear Power in Energy for the 21st Century; Proceedings of the Fifteenth Intersociety Systems, M.S. EI-Genk and M.D. Hoover, eds., Orbit Book Co., Malabar, Energy Conversion Engineering Conference, Vol. 1, AIAA, New York, FL, Vol. 2, pp. 447-453. pp. 143-150.

Ellis, W. 1989. The Space Station Active Thermal Control Technical Chal- Mertesdorf, ST, Pohner, J.A., Herold, L.M., and Busby, M.S. 1987. High lenge. AIAA Paper 89-0073. Power Spacecraft Thermal Management, Technical Report AFWAL- TR-86-2121, TRW Inc., Redondo Beach, CA. Gustafson, E., and Carlson, A.W. 1987. Solar Dynamic Heat Rejection Tech- nology. Task 1: System Concept Development. NASA CR-179618. Oren, J.A. 1982. Flexible Radiator System. NASA CR-171765.

Peterson, G. P. 1987. Thermal Control Systems for Spacecraft Instrumentation. Harwell, W. 1983. The Heat Pipe Thermal Cannister: An Instrument Thermal Journal of Spacecraft and Rockets, vol. 24, no. 1, pp. 7-13. Control System for Space Shuttle. SAE Paper 831124. Ponnappan, R., Beam, J.E., and Mahefkey, E.T. 1984. Conceptual Design of a Juhasz, A.J. 1991. An Overview of the Lewis Research Center CSTI Thermal 1 m Long 'Roll Out Fin' Type Expandable Space Radiator. AIAA Paper Management Program. NASA TM-105310. (Also AIAA Paper 91- 86-1323. 3528. ) Prenger, F.C., and Sullivan, J.A. 1982. Conceptual Designs for 100-Megawatt Juhasz, A.J., and Bloomfield, H.S. 1994. Development of Lightweight Radia- Space Radiators. Report LA-UR-82-3279. Presented at the Symposium tors for Lunar Based Power Systems. NASA TM-106604. (Also pre- on Advanced Compact Reactors Systems, American Nuclear Society sented at the 24th International Conference on Environmental Systems, Winter Meeting, National Academy of Sciences, Washington, D.C., Friedrichshafen, Germany, June 20-23, 1994.) Nov. 15-17.

Juhasz, A.J., and Chubb, D.L. 1991. Design Considerations for Space Radia- Rankin, J.G., Ungar, E.K., and Glenn, S.D. 1989. Development and Integration tors Based on the Liquid Sheet (LSR) Concept. Proceedings of the 26th of the SHARE Payload Bay Flight Experiment. AIChE Symposium Intersociety Energy Conversion Engineering Conference: IECEC-91, Series, Vol. 85, No. 265. Vol. 6, IEEE, LaGrange Park, IL, pp. 48-53. Rovang, R.D. 1988. Advanced Radiator Concepts for SP-100 Space Power Juhasz, A.J., El-Genk, M.S., and Harper, W.B., Jr. 1993. Closed Brayton Cycle Systems. NASA CR-182174. Power System With a High Temperature Pellet Bed Reactor Heat Source for NEP Applications. Proceedings of the 10th Symposium on Space Rovang, R.D., Moriarty, M.P., Ampaya, J.P., Dirling, R.B., Jr., and Hoelzl, R. NuclearPowerand Propulsion Systems, M.S. EI-Genk and M.D. Hoover, 1990. Advanced Radiator Concepts for SP-100 Space Power Systems- eds., American Institute of Physics, New York, pp. 1055-1064. (Also, Phase III Final Report. NASA CR-187170. NASA TM-105933.) Rovang, R.D., Hunt, M.E., and Juhasz A.J. 1994. Testing of a Liquid Metal Juhasz, A.J., and Jones, B.I. 1987. Analysis of Closed Cycle Megawatt Class Carbon-Carbon Heat Pipe. Presented at the 11th Symposium on Space Space Power Systems With Nuclear Reactor Heat Sources. Transactions Nuclear Power and Propulsion. Albuquerque, NM, Jan 10-14. of the Fourth Symposium on Space Nuclear Power Systems, M.S. El-Genk and M.D. Hoover, eds., pp. 423-426. Rutledge, S.K., Hotes, D.L., and Paulsen, P.L. 1989. The Effects of Atomic Oxygen on the Thermal Emittance of High Temperature Radiator Koenig, D.R. 1985. Rotating Film Radiators for Space Applications. Proceed- Surfaces. NASA TM-103224. ings of the 20th Intersociety Energy Conversion Engineering Conference: IECEC'85, Vol. 1, IEEE, LaGrange Park, IL, pp. 439-445. Rutledge, S.K., Forkapa, M., and Cooper, J. 1991. Thermal Emittance Enhancement of Graphite-Copper Composites for High Temperature Koester, J.K., and Juhasz, A.J. 1991. The Telescoping Boom RadiatorConcept Space Based Radiators. AIAA Paper 91-3527. for Multimegawatt Space Power Systems. AIAA Paper 91-3497.

20 Sawko, P.M. 1983. Flexible Thermal Protection Materials. Advances in TPS Weatherston, R.C., and Smith, W.E. 1960. A Method for Heat Rejection for and Structures for Space Transportation Systems, NASA CP-2315, H.N. Space Power Plants. Cornell Aeronautical Laboratory Report, Kelly and J.E. Gardner, eds., pp. 179-192. No. DK-1369—A-1, Ithaca, NY.

Siegel, R., and Howell, J.R. 1980, Thermal Radiation Heat Transfer. Hemi- Webb, B.J., and Antoniak, Z.1. 1988. Rotating Bubble Membrane Radiator for sphere Publishing, New York, pp. 258-261. Space Applications. Proceedings of the 21st Intersociety Energy Conver- sion Engineering Conference: IECEC'86, Vol. 3, ACS, Washington, Song, S.J.; and Louis, J.F. 1988. Rotating Film Radiator for Heat Rejection D.C., pp. 1881.-1885. in Space. Proceedings of the 23rd Intersociety Energy Conversion Engineering Conference: 1988 IECEC, ASME, New York, Vol. I, White, K.A., III. 1988. Moving Belt Radiator Technology Issues. Proceedings pp. 385-390. of the 23rd Intersociety Energy Conversion Engineering Conference: 1988 IECEC, Vol. 1, ASME, New York, pp. 365-371. Teagan, W.P., and Fitzgerald, K. 1984. Preliminary Evaluation of a Liquid Belt Radiator for Space Applications. Final Report. NASA CR-174807. Winter, J.M. 1991. The NASA CSTI High Capacity Power Program. AIAA Paper 91-3629.

21 Form Approved REPORT DOCUMENTATION PAGE OMB No. 0704-0188 Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503. 1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED June 1994 Technical Memorandum 4. TITLE AND SUBTITLE 5. FUNDING NUMBERS Review of Advanced Radiator Technologies for Spacecraft Power Systems and Space Thermal Control

6. AUTHORS) WU-583-02-21 Albert J. Juhasz and George P. Peterson

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT NUMBER National Aeronautics and Space Administration Lewis Research Center E-8263 Cleveland, Ohio 44135-3191

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING AGENCY REPORT NUMBER National Aeronautics and Space Administration Washington, D.C. 20546-0001 NASA TM-4555

11. SUPPLEMENTARY NOTES Albert J. Juhasz, NASA Lewis Research Center; and George P. Peterson, Texas A&M University, College Station, Texas 77843. Responsible person, Albert J. Juhasz, (216) 433-6134.

12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Unclassified - Unlimited Subject Categories 20 and 24

13. ABSTRACT (Maximum 200 words) A two-part overview of progress in space radiator technologies is presented. The first part reviews and compares the innovative heat-rejection system concepts proposed during the past decade, some of which have been developed to the breadboard demonstration stage. Included are space-construc table radiators with heat pipes, variable-surface-area radiators, rotating solid radiators, moving-belt radiators, rotating film radiators, liquid droplet radiators, Curie point radiators, and rotating bubble-membrane radiators. The second part summarizes a multielement project including focused hardware development under the Civil Space Technology Initiative (CSTI) High Capacity Power program carried out by the NASA Lewis Research Center and its contractors to develop lightweight space radiators in support of Space Exploration Initiative (SEI) power systems technology.

14. SUBJECT TERMS 15. NUMBER OF PAGES 24 Space radiators; Space power systems; Space heat rejection 16. PRICE CODE A03 17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT OF REPORT OF THIS PAGE OF ABSTRACT Unclassified Unclassified Unclassified NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. Z39-18 298-102 National Aeronautics and Space Administration Code JTf Washington, D.C. 20546-0001

Official Business Penalty for Private Use $300

POSTMASTER: If Undeliverable — Do Not Return